From Salomon ski equipment to Wilson tennis rackets via Suunto watches, Amer Sports has a dozen or so brands in the sports business. As head of Amer Sports’ Finance Shared Services Centre in Krakow, Poland, Hervé Hachet caters to the operational needs of the group’s teams in Europe, North America and Australia.
Established in 2012, the SSC has focused resources on continuous improvement and has initiated automating tasks from a Microsoft environment. In 2019, the centre signed a partnership agreement with UiPath and has since multiplied its robots. Hervé Hachet shares the experience of his adoption of RPA.
How did you switch to automation?
Improving processes requires tools. The first tool we used consisted of Excel macros which gradually evolved into robots created in a Microsoft environment that became increasingly more sophisticated and was supervised by a local player in Krakow. As the objective of an SSC is to save money, we drew from our teams the employees interested in development, improvements and, of course, Excel formulas to create the core of our automation department.
It took nearly three years to set up the organisation because I didn’t have a budget at the outset. When the number of robots created exceeded sixty, our IT departments considered the challenges of ongoing developments and decided that a second supplier would provide better structuring and increased security to protect sensitive services. They opted for UiPath. Today, the entire mining process is carried out internally, while the development is done externally due to a lack of skills and resources.
What was the most difficult phase in this approach?
Project financing is an important aspect. Although we had no budget at the start, we were able to build a relatively inexpensive base using the Microsoft licences we already had. A company that launches into RPA must nonetheless plan for a significant investment and a relatively long return on this investment. Ideally, processes should be adapted to RPA, but when you have a business like ours that involves several brands, and locations, factories, R&D centres and different entities with dispersed processes nearly all over the world, standardisation is certainly the most difficult project to carry out.
What is the current status of your robot production?
We have about 150 implemented robots. About a hundred are dedicated to classic financial activities, but also to compliance, such as VAT control, for instance. We are also increasingly supplying robots to other divisions of the group that are not involved in finance or directly in our SSC. This is the case for marketing departments for which we have developed a robot capable of identifying the number of times one of our products is named in hundreds of websites and then providing a valuable contextual analysis.
Another robot was designed to provide a representative in the field with more up-to-date information on the availability of stock items, without having to resort to a huge ERP system. Yet another was designed specifically to optimise the responsiveness of IT support teams when a trouble ticket is generated by a user via email: this robot analyses the content of the tickets before dispatching them to different managers, thus taking the place of a person who would distribute trouble tickets to his team.
What ROI have you seen?
Beyond the quantitative there is a qualitative return on investment, particularly because our SSC provides services to other functions within the Amer Sports group. For example, to replace the tedious work of updating sales data every month by a sales representative, we created a robot that performs this task every day, which results in better sales because the more frequently updated data are more relevant and give sales people access to better information. Thanks to RPA, we have automated 15% of our financial processes over a period of 4 years, which is quite interesting for a centre that employs 250 people.
We are far from the 90% automation announced by some market players with tongue rather in cheek. In any event, we calculate the ROI of a robot before manufacturing it. Of course, this ROI must then be confirmed, but we have never had any unpleasant surprises. The cost of manufacturing a robot varies between €5,000 and €10,000, including the cost of licences and the cost of programming the robot. To this must be added the analysis of the processes and the maintenance of the robots, as malfunctions inevitably have to be corrected.
How do you see the use of RPA developing in your activities?
The technological breakthrough that would make robots intelligent enough to replace a person is not yet here. Once again, the greatest difficulty we face is not only to manage to standardise each process, but to make it as rational as possible. And we have to recognise that rationality is not always present in the field of finance. A slightly dramatic example: if you want to automate cash collection, you just have to change the business, abandon the sales made by representatives who go in the shops and digitalise it.
From then on, customers pay for their orders on the Internet and no longer need to be called. We sell with an application and no longer through a customer service. This illustrates the fact that finance is being automated according to business technology disruptions. The digital transformation of companies is underway and it is not going to stop. In our group, 25% of sales are now made via the Internet, whereas 10 years ago they accounted for no more than 5%. To accompany this change, we set up popularisation sessions around RPA early on, with a bottom-up approach that served to gather ideas and suggestions from our employees.
We now prefer a top-down approach to target a particular department or brand and then review all the processes involved. What we may lose in terms of employee support for a project, we gain in terms of speed of execution. All the more so since, in order to identify the best automation opportunities, we are increasingly entrusting dedicated robots with the task of carrying out process mining on the work of volunteer teams. Finally, we are now looking at all our processes from an automation perspective, even if it means having to think out of the box. But adapting long-standing financial processes to make the most of RPA is an approach that requires a lot of expertise.